Why Edge Analytics?

AUTHOR:  Dan Dang

Internet of things (IoT) is an industry growing exponentially as billions of connected devices are expected to be implemented in the coming years. Smart devices are now monitoring temperature, humidity, acceleration, vibration and much more, generating massive amounts of data that need to be analyzed in some way every millisecond. How to analyze this data quickly and efficiently is key for businesses. As a result, enterprises are turning to edge analytics where the analysis occurs at the source of that data, or at the “edge”. There are many advantages from using this approach to analyzing IoT.

Traditionally data is transmitted to a central data center for analysis and decision-making, and then data is transferred back to the source. However, with the enormous increase in IoT devices and data, this method of analytics is not as viable anymore. By utilizing edge analytics, data is processed rapidly within the device and can be acted upon swiftly. This in turn would lower data transmission costs and reduce latency. Moreover, the risk of bottlenecking from pushing data to a centralized node would diminish. Edge analytics therefore provides long-term scalability to the increasing number of connected devices.

The technological requirements of this type of analytics vary in the form of new processing requirements, data flows and network management competencies. Industries considering incorporating edge analytics will require tough servers for stressful situations, along with routers that supply resources in storage, network and virtualized computing. Take for example smart traffic cameras. An anomaly detection algorithm is needed to analyze video stream data at the camera itself in order to prevent a crash. Real-time instantaneous analysis is a must when it comes to traffic safety.

Several industries would benefit immensely from edge analytics, such as health care, oil & gas and energy. Hospital equipment that treats cancer would have boosted uptime and the ability to pinpoint problems as they occur. This results in punctual treatment appointments and fewer doctor visits. Routers at oil rigs can recognize problems with drill bits or other failures and make decisions to resolve those issues in real-time, rather than waiting on delays in data transmissions. Remote monitoring of wind turbines can improve their operational efficiency by communicating with each other, creating a smart grid. The best angles are constantly calculated by wind turbines in order to catch the most wind, thus generating more energy. Ultimately, analyzing data at the edge provides valuable insights and initiates solutions at amazing speeds.

Recently Medium One has built an edge analytics system on Samsung’s ARTIK platform that connects to devices, receives data for local analytics and processes it through workflows. ARTIK is a line of modules designed with IoT in mind. Improved performance, favorable power consumption and memory usage make ARTIK ideal for multiple situations including local processing and analytics. By integrating Medium One workflows, edge analytics is made simple. For more information, visit our website at www.mediumone.com.


Author: Joyce Li